Re: [gmime-devel] Decoding UTF-8 headers

[Andrew Victor <avictor za gmail com>]

h Jeff,

>> It seems that when .NET's System.Net.Mail encodes some headers in
>> UTF-8 format, they cannot be decoded by the GMime library.
>
> Is this Microsoft's System.Net.Mail or Mono's System.Net.Mail or both?

This is with Microsoft's System.Net.Mail  (.NET4)


>> From: =?utf-8?Q?{#D=C3=A8=C3=A9=C2=A3=C3=A5=C3=BD_M$=C3=A1=C3?=
>>
>>  =?utf-8?Q?=AD.=C3=A7=C3=B8m@#}?=
>>
>>  <user domain com>
>>
>> Tested with GMime 2.26, GMime 2.4.14,  Evolution 2.28.3 - only the 1st
>> part of the address is decoded.
>> ie, they display:
>>    "{#Dèé£åý M$á� =?utf-8?Q?=AD. =C3=A7=C3=B8m@ #}?="<user domain com>
>> rather than:
>>    {#Dèé£åý M$áí.çøm@#}<user domain com>
>
> If you initialize GMime with g_mime_init (GMIME_ENABLE_RFC2047_WORKAROUNDS);
> GMime 2.4.x and 2.5.x should both handle the test case you provided. I just
> tested 2.5.8 and 2.4.25 and it works fine. 2.4.14 may be too old, I'm not
> sure.

Thanks.  We're currently using 2.2.26, so I'll look at upgrading to 2.4,


> The problem is that periods must be encoded if they are inside an
> encoded-word token and in your example, that is not the case. The software
> that produced the encoding has a bug.

Ok.

Is it not also an issue that 1 byte of a 2-byte UTF-8 "character"
(=C3=AD) is in one encoded-word, and the other byte in the next
encoded-word (next line)?


Regards,
  Andrew Victor